Skip to main content
. 2018 Jun 26;12:25. doi: 10.3389/fnbot.2018.00025

Figure 4.

Figure 4

Graphical models for concept formation: (A) model for hierarchical concept (Attamimi et al., 2016) constructed with multimodal latent Dirichlet allocations (MLDAs), (B) model for object concept and language acquisition (Nakamura et al., 2014; Nishihara et al., 2017) constructed with MLDAs and speech recognition, and (C) model for location concept and language acquisition (Taniguchi et al., 2017) constructed with simultaneous localization and mapping (SLAM), Gaussian mixture model (GMM), MLDA, and speech recognition.